What is Regression Testing? A Software Testing FAQ - Why? How? When?

What is Regression Testing? In this video you will learn about the legacy definition, and how the term applies to Modern software development processes.

Regression Testing:

- What is it?
- Why do we do it?
- How do we do it?
- When do we do it?
- How did the concept arise?
- Why the phrase Regression Testing is wrong
- How we perform Regression Testing a legacy project
- How we incorporate the concept into Modern software development
- Where does exploratory testing fit in to regression testing?
- How does test automation help with Regression Testing?
- Can Test Driven Development (TDD) help?
- Can Acceptance Test Driven Development (ATDD) help?
- Can Behaviour Driven Development (BDD) help?
- What other Risk mitigation strategies are there?

A short 11 minute video that covers a lot of ground and history of Software Testing.

*************

*************

►► Contact Details:

Read my writing and blogs at:

Follow me on social media:

*************

►► Read my books

★ "Dear Evil Tester" - explore the tester's mind
★ "Java For Testers" - learn to code in Java
★ "Automating and Testing a REST API - a case study"

*************

►►Learn skills from my online training courses

★ Selenium WebDriver With Java
★ Technical Web Testing 101
★ Evil Tester Talks Technical Testing
★ Case Study: Java Desktop Application Technical Training

*************

[desarrolloyarquitecturadigital]

I hope you're doing well! I recently came across your YouTube video on regression testing, and I wanted to express my appreciation for shedding light on this important topic.

With over 20 years of experience in software development and currently serving as a Solutions Architect, I've been exploring regression testing to better understand QA team dynamics and expectations. It's clear that the practice isn't just about trust, but rather a fundamental best practice embraced by QA teams.

While I agree with the importance of mitigating risks, I'm finding it challenging to bridge the gap between theory and practice in our regression testing efforts. Despite our efforts in TDD, Agile, and modular development, comprehensive regression testing remains the norm.

I'm reaching out not to claim expertise, but rather to engage in a constructive conversation. I believe your insights, along with those of your audience, could offer valuable perspectives on rethinking regression testing practices.

Thank you once again for your informative video

[barkgj]

I knew the term from my internship somewhere back in 2001 and up and until now i had no idea what the phrase regression test actually meant, but thanks to you right now i know :) Thanks so much for giving such a clear and logical explanation.

[hakimakbary]

Would you mind making a video about what challenges are there when we want to create our regression testing? How long is a good regression test? what are the best practices for choosing the regression testing candidates? What challenges you (personally) faced in your career about regression testing (no need to give the name of the company or application), how did you resolve them?

[NaeemAkramAndroidiOSApp]

Great! Thanks for sharing. Someone pointed me here in a LInkedIn discussion. I've listened you once, would like to listen one more time to understand deeply.

[shardersharder]

Changes may be of many types. In regression testing you test not just changes related to the changed code, but also more 'indirect' changes: dates change every day (and 0.3% of these are changes in years), environments change (new versions of browsers or OS), rolling out new code may involve changes not directly related to that code (you may have conflicts with other new code, code may get lost in the merge, builds may fail etc). Regression tests should also cover this more uncertain uncertainty, so the risk-analysis is of another type. Asking "What regressions may my new code cause" is not regression testing, but normal functional testing. In regression testing you need to consider the risks _independently_ of the new code. Where may the code, as it was yesterday, it is today and it will be tomorrow, run into problems and, very importantly: where do we most want to avoid problems. (This could be problems that have large impact, ie problem in core functionality, or problems that will more easily go unnoticed for a long time, but still have consequences, ie often problems in edge functionality). For this reason automated testing is useful, as is exploratory testing.

[bjsif]

Thank you for putting this together in way that can be easily understood. I kept confusing this with the statistics term "regression" and I was super confused.

[heshagrade]

Okay, probably I'm gonna be not the first at that, but why is EvilTester not
Thank you, you give an accurate and clear information, it's really nice to listen to you and how you explain the things.

[zourdy]

are you the old version of kent c dodds?

[jivey]

Finally someone who speaks real English making a video on the subject!

[braindeveloperdimensional]

Thank you for this awesome explaination. Can you make a video on manual testing?

[seidemsh]

I was blessed to work with a mature waterfall scenario. Poorly managed software development yields a poor product. Is Agile creating better products today? Had software been built initially with Agile, I think we'd have evolved much slower overall with more buggy sites and apps. Faster is not necessarily better on any level, but that's the song that's being sung far and wide...until the next methodology arises. Then Agile will be viewed as wasteful. LoL

[vicollerbytes]

good info, I dont agree with all, perhaps how it was addressed. Just curious, why you wear long nails, are you guitarist?